diff options
| author | Joseph Myers <joseph@codesourcery.com> | 2014-05-14 12:37:24 +0000 |
|---|---|---|
| committer | Joseph Myers <joseph@codesourcery.com> | 2014-05-14 12:37:24 +0000 |
| commit | 01dbacd22a8d5e0053f8d0cf13a80286b6cfe79b (patch) | |
| tree | 91adc1f6c05cc5dbfb6b3b55b7500fa9dae9fd0e /math/s_cacosf.c | |
| parent | 913d03c864ea2547e97f8d2d30fc71a008d4d103 (diff) | |
| download | glibc-01dbacd22a8d5e0053f8d0cf13a80286b6cfe79b.zip glibc-01dbacd22a8d5e0053f8d0cf13a80286b6cfe79b.tar.gz glibc-01dbacd22a8d5e0053f8d0cf13a80286b6cfe79b.tar.bz2 | |
Fix cacos (+Inf + finite*i) in round-downward mode (bug 16928).
According to C99/C11 Annex G, cacos applied to a value with real part
+Inf and finite imaginary part should produce a result with real part
+0. glibc wrongly produces a result with real part -0 in FE_DOWNWARD
mode. This patch fixes this by checking for zero results in the
relevant case of non-finite arguments (where there should never be a
result with -0 real part), and converts the tests of cacos to
ALL_RM_TEST.
Tested x86_64 and x86 and ulps updated accordingly.
[BZ #16928]
* math/s_cacos.c (__cacos): Ensure zero real part of result from
non-finite arguments is +0.
* math/s_cacosf.c (__cacosf): Likewise.
* math/s_cacosl.c (__cacosl): Likewise.
* math/libm-test.inc (cacos_test): Use ALL_RM_TEST.
* sysdeps/i386/fpu/libm-test-ulps: Update.
* sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
Diffstat (limited to 'math/s_cacosf.c')
| -rw-r--r-- | math/s_cacosf.c | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/math/s_cacosf.c b/math/s_cacosf.c index 9eaeeec..1c9d8b9 100644 --- a/math/s_cacosf.c +++ b/math/s_cacosf.c @@ -34,6 +34,8 @@ __cacosf (__complex__ float x) y = __casinf (x); __real__ res = (float) M_PI_2 - __real__ y; + if (__real__ res == 0.0f) + __real__ res = 0.0f; __imag__ res = -__imag__ y; } else |